A ppt based on predicting prices of houses. Also tells about basics of machine learning and the algorithm used to predict those prices by using regression technique.

1. Housing Price Prediction
-Abhimanyu Dwivedi
-Ashish Gupta
1

2. OUTLINE
 Project Summary
 Technology Used
2
 Tools Used
 How it works ?
 What it does ?

3. Introduction
 Problems faced during buying a house:
1) Buying a house is a stressful thing.
2) Buyers are generally not aware of factors that influence the house
prices.
3) Many problems are faced during buying a house.
4) Hence real estate agents are trusted with the communication
between buyers and sellers as well as laying down a legal contract for
the transfer. This just creates a middle man and increases the cost of
houses.
3

4.  They believe that it depends upon:
1) The square foot area
2) Neighbourhood
3) The no. of bedrooms
 But it depends upon many factors also…..Such as:
1) No. of storeys.
2) Area outside the house.
3) Rooms on one floor.
4

5. Project Summary
 Our project is a machine learning app, based on certain
specifications of your future home it will try to guess the
most accurate price.
 Information such as state, city, area, stores.
5

6. Technology Used
1. Machine Learning 
 Tom Mitchell provides explained: "A computer program is said to learn from
experience E with respect to some class of tasks T and performance measure
P, if its performance at tasks in T, as measured by P, improves with experience
E."
 Example: playing checkers.
 E = the experience of playing many games of checkers
 T = the task of playing checkers.
 P = the probability that the program will win the next game.
6

7. 7
Example: predicting whether the given
object is pen or pencil??
E = the experience of predicting many pens
and pencil.
T = the task of predict pen or pencil
P = the probability that of whether it is a
pen or pencil

8. In general, any machine learning problem can be assigned to one of two broad
classifications:
Supervised learning and Unsupervised learning.
1) Supervised Learning: - In supervised learning, we are given a data set and
already know what our correct output should look like, having the idea
that there is a relationship between the input and the output. Supervised
learning problems are categorized into "regression" and "classification"
problems.
2) In a regression problem, we are trying to predict results within a
continuous output, meaning that we are trying to map input variables to
some continuous function.
8

9. 3) In a classification problem, we are instead trying to predict results in a
discrete output. In other words, we are trying to map input variables into
discrete categories.
• Unsupervised Learning:-Unsupervised learning allows us to approach
problems with little or no idea what our results should look like. We can
derive structure from data where we don't necessarily know the effect of
the variables. We can derive this structure by clustering the data based on
relationships among the variables in the data. With unsupervised learning
there is no feedback based on the prediction results.
9

10. Tool Used
 PYTHON
 TENSORFLOW
 Android Studio
 Anaconda
10

11. HOW IT WORKS ?
 COLLECTING DATA : FIRST STEP WAS TO COLLECT
DATA WE COLLECTED DATA FROM DIFFERENT
SOURCES & MERGED THEM TOGETHER TO FORM OUR
TRAINING DATA SET.
 THEN WE TRAINED THE MODEL USING MACHINE
LEARNIG ALGORITHM WHICH IN THIS CASE IS
MULTIPLE LINEAR REGRESSION.
 BASED ON THE GENERATED GRAPHS WE PREDICT
THE COST OF THE HOUSE
11

12. 12

13. Future Work
 Our model had a low rmse score, but there is still room for
improvement. In a real world scenario, we can use such a model to
predict house prices. This model should check for new data, once in a
month, and incorporate them to expand the dataset and produce
better results.
 We can try out other dimensionality reduction techniques like
Univariate Feature Selection and Recursive feature elimination in the
initial stages.
 We can try out other advanced regression techniques, like Random
Forest and Bayesian Ridge Algorithm, for prediction. Since the data is
highly correlated, we should also try Elastic Net regression technique.
13